Photo by Melinda Gimpel on Unsplash
- Several major property and casualty insurers have submitted regulatory filings in multiple states seeking to add explicit AI liability exclusions to standard commercial policies.
- These exclusions could strip businesses of coverage for losses tied to algorithmic decision errors, AI-generated content disputes, or autonomous system failures.
- The regulatory approval process varies by state, meaning patchwork coverage gaps could emerge well before most policyholders realize it.
- Standalone AI liability insurance products are beginning to enter the market, though policy coverage terms and pricing remain highly inconsistent across carriers.
What Happened
What if the insurance protecting your business quietly stopped covering one of your fastest-growing risk areas — and you only discovered it after a claim was denied?
According to Google News Insurance, citing a report from Insurance Business, a number of large commercial insurers have formally requested regulatory approval to add restrictive language to their standard liability policies — language that would sharply limit or eliminate coverage for losses arising from AI-related incidents. These are not back-room policy conversations. They are official filings submitted to state insurance commissioners, the regulators who must approve any material changes to policy coverage language before insurers can enforce them.
The filings span multiple states and target several commercial lines, including general liability (broad coverage for third-party bodily injury or property damage) and professional liability, also called errors and omissions or E&O (coverage for mistakes in professional services). The driving concern: AI systems introduce liability scenarios that traditional actuarial models — the statistical tools insurers rely on for risk assessment — were never designed to price.
When a physician uses an AI diagnostic tool that misses a critical finding, when a lender's algorithm incorrectly denies a mortgage based on flawed training data, or when an autonomous delivery robot damages a neighbor's property, the question of who pays — and under which policy — has no clean answer in most current commercial contracts. Insurers argue that without clearer exclusions they are underwriting risks they cannot quantify. Consumer advocates counter that businesses and individuals could be left holding the bag for losses from AI systems they didn't design and barely control.
The Insurance Business report notes the pace of these filings is accelerating alongside the rapid enterprise deployment of generative AI, with dockets in at least a dozen state regulatory bodies showing new AI exclusion language as of spring 2026.
Photo by Redd Francisco on Unsplash
Why It Matters for Your Coverage
Here is the real risk embedded in this story: most business owners assume their existing general liability or E&O policy covers whatever goes wrong in their operations. That assumption worked reasonably well in a world of human decisions and physical products. In an AI-integrated world, it may already be wrong — and these new exclusion filings would make it definitively, contractually wrong.
Think of what happened with cyber insurance. For years, standard commercial policies absorbed cyber-related losses under vague "property damage" and "business interruption" language. Then, starting around 2016 and accelerating hard through 2019 and 2020, insurers began carving cyber risks out of standard policies entirely and requiring separate standalone cyber policies. Businesses that missed the shift — or couldn't absorb the added premium — discovered they were uninsured for exactly the events most likely to hit them. AI liability is tracing an eerily familiar arc.
Swiss Re Institute analysts have estimated that AI-related risks could generate more than $750 billion in new insurance premiums globally over the next decade, as the risk assessment landscape matures. But that clarity arrives with a price: as insurers build sharper models around AI exposure, they are pulling back on the broad coverage that previously included these risks by default — and the filing data reflects that shift clearly.
Chart: Estimated number of AI liability exclusion filings submitted to U.S. state insurance regulators by year. The 2026 figure reflects industry analyst projections based on first-quarter regulatory docket activity. Source: industry regulatory tracking data.
The coverage gap is not hypothetical. Reported incidents in recent years include a healthcare provider sued after an AI triage tool was found to have systematically undertreated certain patient groups — with the resulting liability falling outside standard malpractice policy coverage terms. A financial services firm discovered its E&O policy excluded "automated decision-making systems" only when a regulatory fine arrived tied to its AI credit-scoring model. A logistics company whose autonomous routing software caused a vehicle collision found its general liability insurer contesting coverage because no human was actively directing the vehicle at the time.
For small business owners doing any kind of insurance comparison right now, there is one clause to hunt for in renewal documents: language referencing "automated systems," "machine learning," "artificial intelligence," or "algorithmic decisions" in the exclusions schedule. These phrases are appearing where they simply did not exist three or four years ago — and catching them early is where real insurance savings begin, compared to the far costlier path of fighting a denied claim after the fact.
As SmartLegalAI noted in its analysis of how compliance departments are navigating AI partnerships, the regulatory and liability frameworks surrounding enterprise AI are evolving faster than most risk management teams can track — and that gap is precisely what these insurance exclusion filings are designed to address, from the insurer's side of the ledger.
The AI Angle
There is a layered irony running through this entire story: the tools insurers are now seeking to exclude from coverage are also the tools they are deploying to evaluate, price, and process your claims. AI-driven underwriting platforms — from companies like Cytora, Cape Analytics, and Tractable — are already embedded in how major carriers conduct risk assessment for property and commercial lines. Tractable's computer vision AI, for instance, handles auto damage appraisal for hundreds of thousands of claims annually, compressing claims management cycle times from several days down to hours.
That means a single insurer might use one AI model to calculate your premium, deploy a second AI model to adjudicate your claim, and simultaneously file for regulatory approval to exclude coverage for losses your business suffers because of AI. The asymmetry is notable: AI efficiency gains accrue to the insurer's bottom line; AI-generated risks are increasingly being transferred back to the policyholder through exclusionary language.
Emerging insurtech platforms like Vouch — which focuses on tech startup coverage — and Coalition, oriented around cyber and tech liability, are working to close the gap with policies explicitly engineered for AI-forward businesses. Their claims management workflows are more documentation-intensive than standard commercial lines, and their underwriting criteria can shift materially between renewal cycles. But they represent the clearest current path to policy coverage that actually matches how modern businesses operate.
What Should You Do? 3 Action Steps
Ask your broker or agent to walk through any exclusion language added to your policy since the last renewal — specifically anything touching automated systems, machine learning, or AI decision-making. If your current policy coverage was written before 2024, assume it was drafted before AI exclusions became a standard insurer practice and treat it as needing a fresh read. Running a focused insurance comparison across two or three competing carriers at renewal is one of the fastest ways to surface these differences and find insurance savings before you are locked into another cycle.
Just as standalone cyber insurance emerged as a product category after standard policies began excluding cyber risks, dedicated AI liability policies are entering the market. Ask your agent whether your risk profile — particularly if your business uses AI in customer-facing decisions, healthcare services, financial recommendations, or autonomous operations — warrants a separate policy. The additional premium is almost always less damaging than the uncovered exposure you may be carrying right now. A proactive insurance comparison between traditional carriers and insurtech specialists on this specific coverage is worth the conversation.
Effective risk assessment of your own AI exposure starts with knowing what you are actually running. List every AI tool your business uses, what decisions it influences, and whether those decisions could cause harm to a customer, employee, or third party. This is increasingly what underwriters require before they will quote coverage for technology-forward businesses. The cleaner your documentation — including records of human review at decision points — the better your negotiating position at renewal. Good documentation is also the single most consistent factor in lower AI liability premiums across carriers currently writing this coverage.
Frequently Asked Questions
Does using AI tools in my business automatically void my existing general liability coverage?
Not automatically — but it depends entirely on the exclusion language in your specific policy. Some policies contain broad "automated decision-making" exclusions that could apply to AI-assisted operations; others are narrower and target only fully autonomous systems. The safest step is to have a licensed agent review your current policy coverage against the specific AI tools your business actually uses. Do this before your next renewal cycle, not after a claim is filed and the exclusion language is being cited against you.
How does an AI liability exclusion affect my insurance comparison when shopping for commercial coverage?
It should be one of your primary evaluation criteria, especially if your business uses AI in any customer-facing or high-stakes operational context. During an insurance comparison, ask each carrier to disclose their AI exclusion language in writing before you compare premiums. Two carriers offering the same coverage tier may have dramatically different exclusion breadth — making a straight premium comparison meaningless without reading the exclusion schedules side by side.
Will AI-related claims management work differently under a new standalone AI liability policy than under a standard commercial policy?
In most cases, yes. Standalone AI liability policies typically require more detailed incident documentation than standard general liability claims management processes — including logs of AI system outputs, records of human oversight at the time of the incident, and evidence of your internal AI governance practices. Building these documentation habits before a claim occurs is strongly advisable. Ask any prospective insurer what their claims management requirements are for AI-related incidents before you bind coverage — the answer reveals a lot about how they will treat you at claim time.
What types of businesses face the highest risk assessment exposure from AI liability exclusions?
Industries where AI influences consequential decisions about people face the sharpest risk assessment exposure: healthcare (diagnostic support tools, treatment recommendation engines), financial services (credit scoring, fraud flagging), hiring (resume screening, interview analysis tools), and legal services (document review, outcome prediction). But the risk category is widening rapidly. Any business using generative AI in customer communications, product recommendations, or operational automation should evaluate its exposure before relying on existing policy coverage language that predates modern AI deployment.
Are there proven insurance savings strategies for businesses that need AI liability coverage right now?
Yes — and they closely parallel what worked in the early days of cyber insurance. Invest in AI governance documentation (written policies, human oversight procedures, decision audit trails), demonstrate that humans review high-stakes AI outputs before they reach customers, and work with a broker who specializes in technology or professional liability rather than a generalist. Insurers consistently offer better terms and measurable insurance savings to businesses that can show structured AI risk management rather than ad-hoc deployment. The upfront cost of solid documentation is almost always lower than the premium surcharge applied to undocumented AI exposure. Always consult a licensed insurance professional before making changes to your commercial coverage.
Explore Our Network
Disclaimer: This article is for informational purposes only and does not constitute insurance advice. Always consult a licensed insurance agent for personalized guidance.
